552 Gibbs . — On the Development of the 
strobili, that the ovules are in much the same stage on the whole tree at 
one time. Even when last and this year’s strobili are found there is little 
gradation in the stage of either. 
Though constantly looked for, gradation in sizes was rarely obtainable 
on the same tree or even on others in the same locality. 
Professor Macfarlane, on a recent visit to this country, mentioned the 
case of an isolated $ tree of Ginkgo , growing near Philadelphia, on which 
very few fruits matured each year. One year he tried the experiment of 
tying a bunch of <T flowers to the branches, with the result that the tree 
set a large crop of seed. The next year the experiment did not come off ; 
he therefore decided to adhere to the exact date of the first successful 
result, which was again repeated, and that special date to the day has 
invariably proved reliable. The male trees were found to never vary in 
the maturation of the pollen, and by observing that date in the trans- 
portation of the o* flowers the crop of seeds on the isolated ? tree could 
be controlled. 
The experience at Te Aroha suggests a similar condition. Several 
o* trees were quite near and covered with strobili ; the one ? tree also 
bore an abundance of young cones in a lower fertile zone of the young 
shoots in such profusion as to be noticeable, notwithstanding their small 
size, on account of the glaucous bloom covering the entire strobili. Yet it 
was surprising how many were not fertilized. The cones would grow 
vegetatively for a little time and then drop off. 
Possibly only those ovules in a physiologically receptive slate when 
the pollen' is shed can retain it. Stopes and Fuji ( 51 , p. io), in Ginkgo , 
note great uniformity in the development of the different ovules of the 
same tree, or even on different trees, and it has been remarked in the course 
of this work how uniform the stages are for the different species. 
The strobili of P. Totara consist of three (Fig. 54, brsi) to four (Fig. 
55, brsi) bracts, which in the pre-pollination stage are quite unmodified 
(Figs. 54 and 55, brsi). Two are usually fertile (Figs. 55, 58, and 59), but, 
as we have seen before, one ovule is always younger than the other, showing 
that the apparently opposite and decussate arrangement of the bracts is 
merely an ontogenic convenience, not related to the actual organization of 
the plant, and probably induced by the shortness of the axis of the strobilus, 
to be lost on the elongation of the same, or an increase in the number of 
the bracts. 
When three bracts are present (Figs. 54 and 56) the arrangement is 
evidently spiral. Pilger does not take into consideration this contingency, 
where the opposite and decussate theory cannot hold good. For P. Totara 
( 40 , p. 84) he gives the ‘ receptaculum ’ as formed of two or four ‘ scales ’ ; 
whereas, in the youngest stages, two bracts were not seen on the present 
occasion, always three (Figs. 54 and 56) or four (Fig. 55), but the third one 
